We conducted three experiments to assess how nursing frequency affects milk output in early pig lactation. In Exp. 1, nursing behavior of 12 individually penned sows was recorded on d 1 through 3 postpartum. The milk output was measured using the weigh-suckle-weigh method. The spontaneous changes in the nursing frequency between d 1 and 2 were positively correlated with the changes in milk output (rS = .64), and the same was true for changes between d 2 and 3 (rS = .77). In Exp. 2, we forced sows, 7 to 8 d after parturition, to nurse every 35 min (10 sows) or every 70 min (eight sows) for 24 h. Sows nursing at short intervals had more nursings without milk ejection, but they gave 27% more milk and their litters gained 44% more weight during the experimental 24 h than sows nursing at long intervals. Litters nursed at long intervals massaged the udder longer after milk ejection. In Exp. 3, milk output was measured after intervals of 35, 50, and 70 min imposed in varied order on 11 experimental sows in the second week of lactation. In a separate recording of two 50-min nursings, we measured milk output from teats that were, or were not, sucked at the first nursing (i.e., giving within-sow comparison of 50 or 100 min). The milk output after 35, 70, and 100 min did not differ from that after 50 min, although there was a tendency for a slight increase with longer intervals. The results suggest that glands are refilled early after milk ejection and that the increase in milk available with prolonged intervals is only slight. As a result, it is the nursing frequency that plays a crucial role in adjusting the milk output.